GBA Logo horizontal- Facebook- LinkedIn- Email- Pinterest- Twitter- Instagram- YouTube Icon- Navigation Search Icon- Main Search Icon- Video Play Icon- Audio Play Icon- Headphones Icon- Plus Icon- Minus Icon- Check Icon- Print Icon- Picture icon- Single Arrow Icon- Double Arrow Icon- Hamburger Icon- TV Icon- Close Icon- Sorted- Hamburger/Search Icon-
Musings of an Energy Nerd

What’s the Definition of an ‘R-20 Wall’?

It’s surprisingly tricky to determine what the code means by an ‘R-20’ wall

Image 1 of 3
A new guide from the APA explains how a wall insulated with R-19 batts can comply with the code requirement for R-20 insulation. If you follow advanced framing practices, you can obtain a credit for your lower framing factor.
Image Credit: APA

Builders often talk about the R-value of their walls. But if a builder claims to have an R-20 wall, what does that mean?

Building codes commonly include a table listing the minimum prescriptive R-values for walls and ceilings in different climate zones. For example, Table R402.1.1 in the 2012 International Energy Conservation Code (IECC) informs builders that the minimum prescriptive R-value for walls in Climate Zones 3, 4, and 5 is “20 or 13+5.”

This type of table raises many questions. For example, if a builder chooses to comply with the R-20 option, how is R-20 calculated? The code provides some guidance on the issue, but not much. According to a footnote at the bottom of the table, the “first value is cavity insulation, second is continuous insulation or insulated siding, so ‘13+5’ means R-13 cavity insulation plus R-5 continuous insulation or insulated siding.”

Complying with the code usually means installing insulation between the studs

The code language governing the prescriptive R-value requirements has changed in recent years. If you want to know the specific language that is enforced in your jurisdiction, you’ll have to consult your local code book. For example, the prescriptive requirements in the 2009 International Residential Code (IRC) note (in section N1102.1.1), “Computed R-values shall not include an R-value for other building materials or air films.”

A footnote to — equivalent in most respects to the prescriptive table (Table R402.1.1) in the 2012 IECC — notes, “R-19 batts compressed in to nominal 2×6 framing cavity such that the R-value is reduced by R-1 or more shall be marked with the compressed batt R-value in addition to the full thickness R-value.” This footnote is confusing. Who should do the marking? Does this instruction mean that the builder has to mark the insulation in every stud bay? How do you mark an unfaced batt — with spray paint? Or is this footnote directed at insulation manufacturers?

Who knows?

R-19 batts provide about R-18, if you ignore thermal bridging

The footnote about R-19 batts traces its roots to a time when codes required walls in some climate zones to be insulated to R-19. To comply with this requirement, builders usually used 6-inch-thick R-19 batts (originally developed for use on attic floors). Since these batts were labeled “R-19,” many inspectors accepted them — even though, when compressed to 5 1/2 inches, the batts only provided R-18. (For the time being, we’re not even going to address the fact that thermal bridging through the studs reduces the thermal performance of this type of wall to about R-13). This is the issue that code writers were attempting to address with their confusing footnote about “marking” the insulation.

In a later edition of the code (the 2012 IECC), this footnote was changed to read, “When insulation is installed in a cavity which is less than the label or design thickness of the insulation, the installed R-value of the insulation shall not be less than the R-value specified in the table.” In other words, a builder who installs an R-19 batt in a 5 1/2-inch cavity is supposed to know that the (slightly compressed) batt is actually an R-18 batt.

It would be nice if code writers knew how to write English

These examples point to a basic problem: The code is so poorly written that our chief guide through the murky thicket is oral tradition.

“I’m embarrassed by how poorly the code language is chosen,” Joe Lstiburek told me recently. Lstiburek is a principal at Building Science Corporation in Massachusetts who has successfully shepherded several code-change proposals through the Byzantine code-approval process. “It’s horrible, but it is the best we can do. To paraphrase Winston Churchill, the energy code is the worst form of code writing there is except for all of the others. We keep hearing from people who say, ‘We want the code to be simple.’ It isn’t supposed to be a physics or engineering document. Builders just want to know to put in an R-20 batt. They don’t care about thermal bridging. Everybody know what ‘R-20’ means. The only problem is that technological innovation has made this approach obsolete. So how do you fix the way the code is written? It’s almost impossible to fix, because it takes forever, and nobody has an interest in changing the code. The way you get codes changed is that somebody pays you a lot of money to work on getting some aspect of the code changed. Manufacturers pay lobbyists to do that. But nobody has a financial interest in improving bad language in the code. And if you change any aspect of the language, there is almost always somebody who will be offended.”

Not all walls with R-20 insulation are equal

There are several problems with the traditional method used by building inspectors to define an R-20 wall. The most obvious problem is that the R-value of a wood-framed wall is always less that the R-value of the batts installed between the studs.

Another problem is that some walls insulated with R-20 batts perform better than other walls insulated with R-20 batts. For example:

  • Walls with 24-inch-on-center studs have a lower framing factor (and therefore a higher whole-wall R-value) than walls with 16-inch-on-center studs.
  • Walls with single top plates have a lower framing factor (and therefore a higher whole-wall R-value) than walls with double top plates.
  • Walls with insulated headers have a higher whole-wall R-value than walls with uninsulated headers.
  • Walls that are well sealed perform better than walls that have a lot of air leaks.

Using R-19 batts to comply with the R-20 requirement

To address some of these issues, APA — The Engineered Wood Association (a group formerly known as the American Plywood Association) is now arguing that builders who reduce the number of studs in their walls, or who install insulated headers over windows and doors, should be able to use R-19 insulation and still get credit for an “R-20” wall.

Apparently, one of the reasons that APA is taking this position is that the organization is worried about the increasing use of foam sheathing. If the use of foam sheathing causes a reduction in the demand for OSB and plywood, the manufacturers that APA represents will lose business.

The logic behind the APA’s suggested approach is explained in a new APA document, Since the document was reviewed by the International Code Council (ICC) before publication, it includes a stamp of approval from the ICC. The document states, “The assemblies in this publication are deemed to be equivalent to the frame wall assemblies of the prescriptive insulation requirements specified in Table R402.1.1 of the 2009, 2012, and 2013 IECC for those climate zones that require R-20 or R-13+5 insulation.”

However, there’s at least one problem with this explanation: Unfortunately, there are no “frame wall assemblies” listed or described in the prescriptive insulation requirements specified in Table R402.1.1 — just a very vague table with a few obfuscatory footnotes.

The U-factor alternative path

The recommendations made in the APA document are apparently based on one of the four compliance paths offered by the IECC — namely, the U-factor alternative path (found in section 402.1.3 of the 2009 IECC and section N1102.1.2 of the 2009 IRC). The calculations are based a few assumptions; these include the assumption that the typical framed wall has a framing factor of 25%, and that any wall with a lower framing factor can use insulation with a lower R-value than shown in the prescriptive table and can still end up with an equivalent whole-wall R-value.

Another key to the APA’s calculation is its assertion that “the equivalent U-factor for R-20 or R-13+5 wall assemblies will be 0.060 in the 2013 code.” A U-factor of U-0.06 is equivalent to a whole-wall R-value of R-16.6. The APA argues that wall assemblies with a whole-wall R-value of R-16.6 are equivalent to walls that meet the R-20 prescriptive requirement.

So, what if a builder wants to use R-19 batts, which are cheaper than R-20 batts? Here’s what you do, according to APA: you use advanced framing methods, and give yourself credit for having a lower framing factor than other walls. You also give yourself credit for air films, for vinyl siding, for 7/16-inch OSB, and for drywall. (We’ll leave aside, for the time being, the fact that the 2009 IECC specifies in section N1102.1.1 that “Computed R-values shall not include an R-value for other building materials or air films.” Presumably, that limitation applies only to builders following the prescriptive path, not to builders following the U-factor alternative path. But wouldn’t it be nice if the code were written clearly enough that builders didn’t have to guess what the code means?)

Here’s how it adds up: Vinyl siding is R-0.59; OSB is R-0.62; fiberglass batts are R-18; two air films and 1/2-inch drywall are R-1.38. Then you adjust your calculation to take into account the framing lumber. You assume a framing factor of 20% for advanced framing, and assume that the framing has an R-value of R-6.88. Using the two-dimensional assembly U-factor calculation method outlined in ASHRAE Fundamentals, the wall assembly is determined to have a U-factor of 0.06 and a whole-wall R-value of 16.67.

R-19 batts are significantly cheaper than R-20 batts

What’s the point of this exercise? For production builders worried about every penny, this approach allows you to downgrade from R-20 batts to less expensive R-19 batts. And for the APA, it provides an inexpensive option for some builders who may be tempted to install rigid foam sheathing — a threat to the market for OSB and plywood. As the APA points out, “All of the assemblies shown in this guide include the use of continuous minimum 7/16 Performance Category wood structural panel wall sheathing.”

The APA guide includes several more suggested wall assemblies that use less insulation than required by the prescriptive table in the code, and yet still comply with the IECC.

Code compliance options

If the complexity of the above arguments leaves you shaking your head, it’s worth pointing out that there are several more ways to comply with the energy code.

Most versions of provide four compliance paths:

  • The prescriptive path, which requires builders to follow the simple table that lists minimum R-values for walls, ceilings, and floors in different climate zones;
  • The U-factor alternative — the approach used by APA to argue in favor of certain wall assemblies that use R-19 batts;
  • The Total UA alternative; and
  • The simulated performance compliance path.

The “Total UA” alternative path

The “Total UA” alternative compliance path is described in section R402.1.4 of the 2012 IECC:

“R402.1.4 Total UA alternative. If the total building thermal envelope UA (sum of U-factor times assembly area) is less than or equal to the total UA resulting from using the U-factors in Table R402.1.3 (multiplied by the same assembly area as in the proposed building), the building shall be considered in compliance with Table R402.1.1. The UA calculation shall be done using a method consistent with the ASHRAE Handbook of Fundamentals and shall include the thermal bridging effects of framing materials. The SHGC requirements shall be met in addition to UA compliance.”

Builders who choose the Total UA Alternative path can choose insulation thickness and window U-factors that deviate from those in the prescriptive table, as long as the total building thermal envelope UA (the sum of each assembly U-factor times each relevant area) isn’t more than the total UA resulting from using the U-factors in the prescriptive table. If builders want to follow this path, they usually hire an energy consultant to perform the necessary calculations.

Simulated performance path

The simulated performance path uses computer software to calculate an energy budget for a proposed home design. Confusingly, the code calls this compliance path “performance-based compliance,” even though it is based not on performance, but on computer modeling.

This path is described in section 405.3 of the 2009 IECC: “405.3 Performance-based compliance. Compliance based on simulated energy performance requires that a proposed residence (proposed design) be shown to have an annual energy cost that is less than or equal to the annual energy cost of the standard reference design. Energy prices shall be taken from a source approved by the code official, such as the Department of Energy, Energy Information Administration’s State Energy Price and Expenditure Report. Code officials shall be permitted to require time-of-use pricing in energy cost calculations.”

Builders following this path must use energy modeling software to show that a proposed house design has an annual energy budget (in dollars) less than or equal to that of a similar house (known as the “standard reference design”) that complies with the code’s prescriptive requirements.

What does the code say about slab insulation?

Insulation requirements for slab-on-grade floors can be found in section R402.2.9 of the 2012 IECC. (Similar provisions in the 2009 IRC can be found in section N1102.2.8.) According to that section, “Slab-on-grade floors with a floor surface less than 12 inches below grade” are required to be insulated, and the insulation should comply with the requirements of the prescriptive table (that is, Table R402.1.1 in the IECC, or Table N1102.1 in the IRC).

The code is silent concerning whether above-grade slabs or slabs that are more than 12 inches below grade need to be insulated. The vast majority of slab-on-grade floors are actually above grade rather than below grade; the code implies that above-grade slabs don’t need to be insulated.

That interpretation doesn’t correspond with oral tradition, but that is what the code says.

Confused yet?

Most of the compliance methods described above are flawed. Some — especially the prescriptive method laid out in the table that tells builders to install insulation between studs — are deeply flawed. Others — including the simulated performance path — are only a little bit flawed.

Joe Lsiburek’s pessimistic conclusion — that badly written building codes are “almost impossible to fix” — is probably correct. For years to come, we’re going to be stuck with the International Residential Code and the International Energy Conservation Code, flaws and all. So we might as well learn as much as we can about our different compliance options.

Martin Holladay’s previous blog: “EMFs and Human Health.”


  1. Matt Dirksen | | #1

    Oh I Wish...
    Thanks for your post. I am very glad to know "it's not just me". It seems that EVERY TIME I comb through the IRC/IECC I end up with a terrible hangover.

    Given that more than just written words are "admissible" nowadays, I surely wish the code was an interactive "living model". Something where I could go on-line, plug in my project's zip code, and click at different pieces of a beautiful model which showed me what I could (and could not do.)

    Too much to ask for, I guess.

  2. User avater
    Armando Cobo | | #2

    Code Books? What are those...
    When I teach classes, I always make a point to ask how many people owns a code book, never mind that in most municipalities is a combination of several code books. Anyways...It's usually ZERO, but now and then I'm surprised when one or two SAY they do. That applies to Builders, Architects and Designers... very few Subs take these classes.

  3. Malcolm Taylor | | #3

    Point taken - but I think stretched a bit. I've never met or heard of an architect who didn't own a code book. They may not have been familiar with all its requirements but they have one.

  4. User avater
    Armando Cobo | | #4

    Sorry but it's the truth
    Malcom, all the comments I make in any blog or website are 100% my experiences, no BS, lies or stretches. I just taught a class in Santa Fe with 20+ Builders and Architects, and not a single one person in that class owned a code book. Many of them read this website and they can tell if I'm "stretching" it. Same in TX, CO, KS, OK and in almost every class I've taught. Just because this is not your experience, you should not call me less than truthful.

  5. Malcolm Taylor | | #5

    I didn't mean to imply you are liar. I simply don't understand how this can be the case. You can not function as an architect without constant access to the building code. Even for small practices only working on individual houses I don't see how it can be done. While they may have told you this for some reason, I'd be interested to see if any come forward to this website and support your assertion.

  6. Matt Dirksen | | #6

    Own vs. access
    One thing is owning a copy of the code, another thing is knowing how to quickly access information in it.

    We have a full pdf version of the 2009 IRC, but don't actually "own" the 2012. One reason in particular:
    and our local amendments:

    Many of my books sit on the shelf, collecting dust. This is due large in-part to the my ability to access current information much more quickly on-line than finding the book, chapter, paragraph, and keyword. By the time a book is actually published, a certain amount of it is conceivably already outdated.

    Of course this method is no substitute for simply "knowing' the right answer, but I can't always trust what I "know" is actually to code anymore, due to how quickly it all changes, especially in my particular state (the only one in the Country which automatically adopts the latest code as soon as it comes out).

  7. Douglas Horgan | | #7

    IECC does have language about slab insulation
    There is a free copy of the 2012 IECC available here:

    This section explains what is meant by slab insulation "depth". Once you get to this section it's relatively clear:

    R402.2.9 Slab-on-grade floors.
    Slab-on-grade floors with a floor surface less than 12 inches (305 mm) below grade shall be insulated in accordance with Table R402.1.1. The insulation shall extend downward from the top of the slab on the outside or inside of the foundation wall. Insulation located below grade shall be extended the distance provided in Table R402.1.1 by any combination of vertical insulation, insulation extending under the slab or insulation extending out from the building. Insulation extending away from the building shall be protected by pavement or by a minimum of 10 inches (254 mm) of soil. The top edge of the insulation installed between the exterior wall and the edge of the interior slab shall be permitted to be cut at a 45-degree (0.79 rad) angle away from the exterior wall.

  8. User avater GBA Editor
    Martin Holladay | | #8

    Response to Douglas Horgan
    Thanks very much for providing that code reference. It's a section that I somehow missed. I'm grateful for the reference, and I have edited my article for accuracy.

    The most curious aspect of section R402.2.9 of the 2012 IECC is that it does not require any insulation for above-grade slabs. The only slabs that are required to be insulated are “Slab-on-grade floors with a floor surface less than 12 inches below grade” -- a rare type of construction.

    Most concrete slabs are either above grade (usually between 4 inches and 8 inches above grade), or are more than 12 inches below grade (for example, basement slabs).

  9. Aj Builder, Upstate NY Zone 6a | | #9

    iECC is NOT the code every where
    NYS Code.... Is.... NYS Code.

    We modify IECC code and implement codes years later than published.

    And if code is a builders or Arch's problem then they sure must have greater issues.

  10. Malcolm Taylor | | #10

    Response to Martin,
    I think you are reading it a bit too narrowly. Isn't the distinction that provision is trying to address that between slabs more than 12" below grade and all other slabs above that height?

  11. User avater GBA Editor
    Martin Holladay | | #11

    Response to Malcolm Taylor
    Yes, I think you've put your finger on it -- it's probably another example of the inability of code writers to write English.

    Let's see -- is there a difference between slabs that are "below grade" and slabs that are "above grade"? Why, yes -- I think there is.

    But the code writers never mentioned what to do with slabs that are above grade.

    Are these two sentences equivalent? You decide.

    1. My proposed version: "Slab-on-grade floors that with a floor surface that is above grade or that is less than 12 inches (305 mm) below grade shall be insulated in accordance with Table R402.1.1."

    2. The existing language: "Slab-on-grade floors with a floor surface less than 12 inches (305 mm) below grade shall be insulated in accordance with Table R402.1.1."

  12. Malcolm Taylor | | #12

    Your wording would help a lot.
    Obscure code language can have much larger consequences in rural areas where most of the enforcement is done by inspectors on site when things are already built, rather than plan checking before construction starts. It also undermines client's faith in their designer when you can't guarantee that what you draw will pass inspection.

  13. User avater GBA Editor
    Martin Holladay | | #13

    Response to Malcolm Taylor
    If only the International Code Council would hire a few unemployed English majors, we could clarify the building code and lower the unemployment rate for liberal arts graduates.


  14. Malcolm Taylor | | #14

    Send a few our up here, ours is just as bad.
    I think part of the problem is that code writers write for code writers. I changed the way I label drawings and write specifications when I realized that architects write for architects, rather than the builders and subs that have to interpret them. Much of the jargon can be replaced by plain english equivalents, and a lot of it is included just to cover our asses.

  15. Marcus Sheffer | | #15

    90.1 Appendix A
    Our work is mostly commercial and we use ASHRAE 90.1 Appendix A to determine whole assembly U-values. Not perfect but it is a pretty organized and understandable way to estimate the whole assembly performance.

  16. Tony Ferrera | | #16

    Hangover starting
    2. The existing language: "Slab-on-grade floors with a floor surface less than 12 inches (305 mm) below grade shall be insulated in accordance with Table R402.1.1."

    I confused by your interpretation Martin. The code seems to ask that slabs near or above grade be insulated. Makes sense. That surface is colder (winter) or warmer (summer) than anything below 12". To my mind a basement slab 4 or more feet below grade would need less or no insulation as that ground surface is much more moderate in temperature through the seasons than the ground surfaces closer to grade. This is the principle of geothermal energy is it not?

  17. User avater GBA Editor
    Martin Holladay | | #17

    Response to Tony Ferrera
    I'm not questioning the logic that calls for less insulation for slabs that are deeply below grade than for slabs that are near grade. That makes sense.

    What I am questioning is the confusing language of the code, which discusses the requirements for slabs that are "below grade," but which is silent on what should be done with slabs that are above grade.

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |